Original article

Textile Research Journal

2014, Vol. 84(5) 513–526

A new thermocouple technique for the ! The Author(s) 2014

Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
precise measurement of in-plane capillary DOI: 10.1177/0040517513503729
trj.sagepub.com
water flow within fabrics

Chunhong Zhu1 and Masayuki Takatera2

Abstract
A new method was proposed for the measurement of in-plane capillary water flow within fabrics automatically. The
method was based on the use of thermocouples to measure temperature changes as the fabric absorbed water. The
wicking length of three different woven fabrics and two knitted fabrics in the warp, weft and 45 bias directions was
measured using this technique. The wicking length was found to have a linear relationship with the square root of time, in
agreement with the Washburn equation. Comparing the results with the data acquired using the horizontal Byreck
method, the thermocouple technique was found to be suitable for the precise measurement of in-plane capillary water
flow through fabrics automatically. Moreover, based on this method, the feasibility of estimating the water contents of
fabrics from temperature differences generated during the flow process was also investigated. We found that it was
possible to estimate the water content of a fabric, provided the moisture level is below the critical water content at
equilibrium. This can be used to estimate the water content of fabric when it is worn.

Keywords
Thermocouple, temperature, capillary water flow, wicking, woven fabric, water content

The transport of liquids through textiles plays an
important role in their applications. In general, wetting
is the displacement of a solid–air interface by a solid–
liquid interface. The motion of a liquid through a
fibrous assembly, such as yarn or fabric, may be
driven by external forces or solely by capillary action.
The spontaneous transport of a liquid driven into a
porous system by capillary forces is termed ‘‘wicking’’.
Because capillary forces are caused by wetting, wicking
is essentially the outcome of spontaneous wetting in a
capillary system.1–3 Wicking takes place only in wet
fabrics or when fabrics come into contact with water,
and the contact angle determines their wicking behav-
ior. A lower contact angle results in higher wicking
rates.4
Wicking in fabrics has been widely researched both
in academia and industry because it plays an important
role in many textile applications, including the design
and use of sportswear and industrial uniforms, as well
as in various processes, such as dyeing, finishing and
filtration. In applications such as active sportswear,
exercise garments, work clothing, intimate apparel
and footwear, the concept of moisture management is
utilized to prevent or minimize the collection of liquids
on the skin of the wearer due to perspiration. This is
done by quickly wicking or diffusing the liquid through
a hydrophobic fiber inner layer to an outer hydrophilic
layer and then evaporating it to the atmosphere.5
A variety of techniques are used to experimentally
study wicking in fabrics, the simplest of which consists
of following the movement of liquid throughout the
textile structure by using colored water, camera or
image analysis techniques and so on.2,3,6–15 This can
be performed when the fabric is perpendicular to, and
partly immersed in, a liquid bath or when a drop of
1
Interdisciplinary Graduate School of Science and Technology, Shinshu
University, Japan
2
Faculty of Textile Science and Technology, Shinshu University, Japan
Corresponding author:
Masayuki Takatera, Shinshu University, 3-15-1 Tokida Ueda-shi Nagano-
ken 386-8567, Japan.
Email: takatera@shinshu-u.ac.jp

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

514
liquid is applied to a fabric or yarn and subsequently
spreads. The Byreck method, or DIN 5392411,15,16 is a
conventional method of assessing the water absorbency
of fabrics. This method is to use a preconditioned strip
of the test fabric, suspended vertically with its lower
end immersed in a reservoir of distilled water, to
which may be added a dye (of a type known not to
affect the wicking behavior) for tracking the movement
of water. After a fixed time has elapsed, the height
reached by the water in the fabric above the water
level in the reservoir is measured. However, when
applying this method, it is important to take into con-
sideration the influence of the dye applied to color the
liquid on the absorption properties of the fabric, as well
as variances in the apparent liquid flow front caused by
different observational positions and the effects of grav-
ity.17 Although present-day photographic and image
analysis technologies can easily track the surface capil-
lary flow within a fabric, these technologies are not
always suitable in the case of multilayers, thick fabrics
or composite materials. An alternate technique is to
weigh the fabric or yarn during capillary wicking.18,19
However, this method is too sensitive to use with yarns,
since in these textiles the wetting force exceeds the
effects of capillary forces,2,9 and it will not show the
effect of the capillary exactly. Yet another approach
consists of measuring the wicking length by applying
liquid-sensitive sensors, which track electrical capaci-
tance, humidity and other factors, at regular intervals
along the fabric.17,20,21 This approach also has some
associated challenges: the sensors are expensive, and
are not suitable for testing fabrics while they are
being worn, and may have limited durability.
In the work reported herein, a new method of experi-
mentally measuring capillary liquid flow by using
thermocouple array was developed, based on previous
studies of temperature variations in fabrics during the
wetting and drying processes.22 As each thermocouple
is physically very small and measures temperature over
a correspondingly small area, the thermocouple array
has numerous potential applications, not only in study-
ing liquid water transport but also in following textile
water content in real time. By employing removable,
Figure 1. Capillary action.
Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015
Textile Research Journal 84(5)
flexible thermocouples, this method can be used to
determine the wicking length of liquid moving through
fabrics over time without the requirement to add dyes
to the liquid. It is also adaptable to many different tex-
tile structures, including woven and nonwoven fabrics,
as well as yarns. Moreover, this method can be applied
to thick fabrics simply by modifying the thermocouple
measurement points. The aim of this work was there-
fore to design and test a new experimental technique for
measuring capillary liquid flow, and to investigate the
viability of applying this method to estimate the water
contents of fabrics.
Basic theory
The kinetics of wicking in textiles is critical in many
applications and is often investigated by fitting the
experimental data to the well-known Washburn equa-
tion.23 Over short spans of time, when the effects of
gravity are neglected and when the extent of liquid
flow is significantly less than the maximum equilibrium
length, the Washburn equation24,25 states that the capil-
lary water flow rate depends on the rheological proper-
ties of the liquid and the geometry of the capillary
spaces and the surface tension between liquid and
vapor, as shown in Figure 1.
The Poiseuille’s equation is given by Equation (1):
dL R2 P
¼ ð1Þ
dt 8L
where L is the wicking length, t is time, R is the equiva-
lent radius of the capillary space and  is the liquid
viscosity. DP is the capillary pressure difference,
which can be obtained by Equation (2):
2 cos 
P ¼  gh0 ð2Þ
R
where  is the liquid–vapor surface tension,  is the
contact angle of the solid–liquid system,  is the
volume density of fluid, g is the gravitational acceler-
ation and h0 is the distance between the water level in
Zhu and Takatera
the fluid reservoir and fabric laid horizontally (the first
measurement point).
Integrating Equation (1) with initial condition of
L ¼ 0 at t ¼ 0, and substituting Equation (2) for DP,
Equation (3) can be obtained:
 
2 R2 2 cos 
L ¼  gh0 t:
4 R
As h0 is very slight, in order to simplify the calcula-
tion, the horizontal wicking distance of the water is
therefore given by Equation (4) below:
L ¼ ðR cos =2Þ1=2 t1=2 :
According to Equation (4), a plot of L versus t1/2
should be linear and pass through the origin. The
slope of the line, referred to as the wicking coefficient
Wc, is given by Equation (5) and is determined by fit-
ting the experimental data to Equation (4):
Wc ¼ ðR cos =2Þ1=2 :
Experimental details
The experimental trials were divided into two parts.
The initial trials verified the applicability of the new
experimental technique of using a thermocouple array
Figure 2. Thermocouple test apparatus: (a) schematic diagram of the thermocouple method; (b) block diagram of the overall
system.
Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015
515
to measure capillary liquid flow, while subsequent work
investigated the feasibility of employing this method to
estimate the equilibrium water contents of wetted
fabrics.
Experiments for measuring capillary liquid flow
ð3Þ Test apparatus. Figure 2 presents diagrams of the test
apparatus used to measure capillary liquid flow
through the fabrics. The apparatus consisted of an
array of nine thermocouple measurement points set
10 mm apart and sitting on foam polystyrene for heat
insulation. The thermocouples were Type T, made from
ð4Þ copper and constantan element wires of 0.1 mm in
diameter.22 The standard temperature range is from
200 C to 350 C, and the accuracy of the thermo-
couple is within 0.5 C. The temperature data were col-
lected and recorded at a sampling rate of 5 s–1, or 5 data
points per second, automatically.
For the woven fabric, the wetted end of the fabric
strip was clipped to a weight that produced a tension of
ð5Þ 5.0 gf/cm.17 This functioned both to keep the fabric end
immersed in the 20  1 C ion-exchanged water and to
ensure the rest of the fabric strip remained in contact
with the sensor array. The absolute minimum amount
of tension necessary was applied, since excess tension
can strain the fabric and alter its wicking performance.
For the knit fabric, because it is stretchy, a slight ten-
sion may cause the stretch of fabric; the tension for it
516
was only its own weight. The other end of the fabric
was fixed so as to keep it taut throughout the experi-
mental trial. The water level in the fluid reservoir was
maintained at a constant height by perforating a hole
on the reservoir to keep the water level the same before
each experimental trial and suppressing water evapor-
ation during the experiment. As the fabric absorbed
water, changes in temperature were measured by the
thermocouple array.
Principle of the measurement. The temperature of the
fabric at the liquid flow front will change during the
process of water absorption because of heat transfer-
ence associated with both wetting and evaporation. As
the fabric initially absorbs water, wetting heat is gener-
ated and the local temperature temporarily increases,
after which it begins to drop.22 These temperature vari-
ations are measured by the thermocouple at each meas-
urement point and are relayed to and recorded by a
computer.
Figure 3 shows the relationship between temperature
and time during the liquid flow process of a cotton/
polyester fabric. Because of absorption heat, the
fabric temperature increases, and then it goes down
because evaporation is larger than absorption until it
reaches about the wet bulb temperature.22
Since the fabric temperature will rise because of the
absorption of liquid, if the temperature T at time t
increases 0.5 C in less than 10 seconds, then t is defined
as the arrival time, as shown in Figure 3. The 10 second
time scale was determined on the basis of measurement
error and the stability of measured temperature. If the
Figure 3. Temperature variations over time in a cotton/poly-
ester fabric.
Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015
Textile Research Journal 84(5)
temperature at time t increases 0.5 C at tT + 0.5 in which
the interval is more than 10 seconds, then the arrival
time is defined as Equation (6). However, if the wetting
heat is minimal compared with heat loss via evapor-
ation, then the temperature at the flow front will
decrease. In this case, the arrival time of the flow
front is defined as the time at which the temperature
is 0.5 C greater than the temperature at time t + 10 (s),
as shown in Equation (7):
tn ¼ tTþ0:5 C  10s: ð6Þ
tn ¼ tT0:5 C  10s: ð7Þ
Here T ( C) is the temperature at the arrival time t (s).
As nine measurement points are set in this experiment,
n refers to 1–9.
As shown in Figure 3, the time required for liquid
flow between any two measurement points n and n + 1
can be calculated according to Equation (8):
tn!nþ1 ¼ tnþ1  tn , n ¼ 1, 2, . . . , 8: ð8Þ
The overall time (Timen) from the zero of the flow
time to the point n + 1 can be calculated according to
Equation (9):
Timen ¼ t1!2 þ t2!3 þ    þ tn!nþ1 : ð9Þ
Test samples. Three kinds of plain woven fabrics and
two kinds of knitted fabrics were used in this work,
with the specifications shown in Table 1.
Prior to testing, fabric samples were washed with
soap for 35 minutes, then rinsed with water for 50 min-
utes to prevent soap remaining on the fabric samples,
and dried by air at constant environmental condition of
20  1 C and a constant relative humidity (RH) of
65  2%. Then test specimens (25 cm  3 cm) were care-
fully cut from each fabric at varying locations along the
warp, weft and 45 bias directions. Experiments were
performed at a constant temperature of 20  1 C and a
constant RH of 65  2%. Five samples for each direc-
tion of each fabric were tested, and the averages and
standard deviations were determined. A diagram of a
typical fabric strip sample as used in these trials is pre-
sented in Figure 4. The portion of the fabric that is bent
towards the water reservoir is 5 cm long, and 4.7 and
4.8 cm lengths of it are immersed in the ion-exchanged
water, while the horizontal section is 9 cm and is
marked at 1 cm intervals.
To assess the validity of our new measurement tech-
nique, we simultaneously applied the horizontal Byreck
method26 to measure liquid flow through the same fab-
rics. In the following discussions, we refer to this
Zhu and Takatera 517

Table 1. Specifications of test samples

Sample name C C/P-1 C/P-2 C-k P-k

Fiber content 100% Cotton 35% Cotton 35% Cotton 100% Cotton 100% Polyester
65% Polyester 65% Polyester
Fabric type Woven Woven Woven Weft knit Weft knit
Yarn count (tex) Warp 28.0 12.5 13.5 24 21
Weft 28.0 14.5 13.5
Fabric density Warp/Wale 27 56 56 27 32.5
(cm–1) Weft/Course 22 28 28 18.5 22
Mass per unit 153.52 114.60 111.02 176 165
area (g/m2)
Thickness (mm) 0.745 0.442 0.569 1.02 1.03
Porosity2 86.62 82.53 86.41 88.80 88.39

0.001 g. The temperature was measured and recorded

Figure 4. Diagram of a typical fabric sample as used in water
flow trials.
protocol as the Byreck method and the new technique
simply as the thermocouple method.
Estimating water content based on temperature
differences in fabrics
Initially, experimental trials were carried out with
the aim of investigating the relationship between
fabric temperature and water content, using square
sections of fabric. Moreover, the relationship
between water content and difference of fabric and
room temperatures will be obtained, which will be
used for estimating water content from the tempera-
ture difference.
Prior to each trial, test specimens were wetted by
soaking in distilled water for 24 hours and squeezed
to make sure no water dripped from the fabric. Then
the fabrics were laid on the foam polystyrene placed on
the balance, which was set to zero at the beginning of
the experiment. The mass and temperature of each
wetted specimen were measured and recorded during
the subsequent drying process. The continuous mass
was measured and recorded by the GF-600 balance
(A&D Company, Limited), with the resolution of
by three thermocouples with the measurement points
set between two fabrics.
We next carried out experimental trials designed to
investigate the relationship between temperature and
water content at different locations along the fabrics
when water flow had reached equilibrium.
Measurements of the equilibrium wicking lengths
were performed using the same apparatus as shown in
Figure 2, applying thermocouple measurement points as
summarized in Figure 5. The first measurement point
was located immediately adjacent to the water reservoir,
followed by an additional 10 measurement points every
3 cm and then 11 measurement points at 1 cm intervals.
Fabric specimens 55 cm in length and 3 cm wide were
used and the equilibrium wicking length was determined
by dipping a 4.5 cm end portion of the sample in a water
reservoir for 12 hours.23 Temperatures along the fabric
at the thermocouple measurement points were recorded
during this time period, and the temperature of each
measurement point was assumed to be the average tem-
perature of the segment to which the measurement point
was attached. The temperature difference of the fabric at
each segment was calculated by subtracting the mea-
sured temperature at each measurement point from
room temperature.
When the 12 hours wicking time had elapsed, every
1 cm segment of the fabric centered on each of the
thermocouple measurement points was cut off and
weighed to calculate the water content. The water con-
tent of each section of fabric (M) was calculated using
Equation (10):
W1  W2
M¼  100%: ð10Þ
W2
Here, W1 is the mass of the wet fabric segment
immediately after being removed from the test

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

518 Textile Research Journal 84(5)

Figure 5. A typical fabric strip sample as used in equilibrium water flow trials, showing thermocouple locations.

Figure 6. Comparison of the results of thermocouple and Byreck measurements for cotton fabric in the warp direction:
(a) relationship between the wicking length and time; (b) relationship between wicking length and square root of time.

specimen, and W2 is the mass of the same fabric seg-
ment after drying.
A 100% cotton fabric (C) was used in these experi-
ments and six trials were performed along the weft dir-
ection at a constant temperature of 20  1 C and
65  2% RH.
Results and discussion
The feasibility of the thermocouple method
Figure 6(a) summarizes the results of measurements of
wicking length over time during wetting of cotton
fabric in the warp direction, as measured by both the
thermocouple array and the Byreck method.
Figure 6(b) plots the same results but against the
square root of time, which Equation (4) suggests
should produce a straight line.
It is evident that both methods generated data show-
ing essentially the same trends: as wetting time increases,
so does wicking length. Equation (4) predicts that the
wicking length will be zero at time zero, although, as
shown in Figure 2, there is a finite distance between
the water surface and the nearest measurement point,
and so these plots do not go through the origin.
Figure 6(b) shows that the plots of wicking length
versus square root of time are highly linear, with coeffi-
cients of determination above 0.99, in agreement with
Equation (4). The wicking coefficient of the cotton fabric
can be obtained from these data, and values of 0.3099
and 0.3084 are calculated from the thermocouple and
Byreck method data, respectively. There is evidently a

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

Zhu and Takatera 519

slight difference between the two methods such that, at
the same wicking length, the time measured by the
thermocouple method is slightly shorter than that deter-
mined by the Byreck method. This is mainly because the
Byreck method requires that the water front be deter-
mined by visual analysis of color change, and so there is
an associated time lag between the actual and the
observed water front. The thermocouple method,
being automated, with a high frequency of data collec-
tion, is therefore more precise than the Byreck method.
Figures 7 and 8 show the results for cotton fabric (C) in
the weft and 45 bias direction, respectively.

Figure 7. Comparison of the results of the two measurements for cotton fabric in the weft direction: (a) relationship between the
wicking length and time; (b) relationship between wicking length and square root of time.

Figure 8. Comparison of the results of the two measurements for cotton fabric in the 45 bias direction: (a) relationship between
the wicking length and time; (b) relationship between wicking length and square root of time.

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

520 Textile Research Journal 84(5)

Liquid flow throughout the other fabrics (C/P-1,
C/P-2 and C-k) was also measured by both the thermo-
couple array and Byreck techniques. The results are
shown in Figures 9–17 and very similar results were
obtained. Because of the hydrophobicity of polyester
knit fabric (P-k), the wicking length is almost zero,
which cannot be measured and recorded by both tech-
niques. Therefore, the results of polyester fabric cannot
be shown.
Table 2 lists the wicking coefficients determined for
fabric samples in the warp, weft and 45 bias directions,
as measured by the thermocouple and Byreck methods.

Figure 9. Comparison of the results of the two measurements for cotton/polyester fabric (C/P-1) in the warp direction: (a) rela-
tionship between the wicking length and time; (b) relationship between wicking length and square root of time.

Figure 10. Comparison of the results of the two measurements for cotton/polyester fabric (C/P-1) in the weft direction:
(a) relationship between the wicking length and time; (b) relationship between wicking length and square root of time.

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

Zhu and Takatera 521

Figure 11. Comparison of the results of the two measurements for cotton/polyester fabric (C/P-1) in the 45 bias direction:
(a) relationship between the wicking length and time; (b) relationship between wicking length and square root of time.

Figure 12. Comparison of the results of the two measurements for cotton/polyester fabric (C/P-2) in the warp direction:
(a) relationship between the wicking length and time; (b) relationship between wicking length and square root of time.

The relative differences between the results determined Here, Wc(T) is the wicking coefficient determined by
by the two methods Wc are also provided, as calcu- the thermocouple method and Wc(B) is the wicking
lated according to Equation (11): coefficient resulting from the Byreck method. As the
Wc values in Table 2 demonstrate, the differences
  between both methods are small.
WcðTÞ  WcðBÞ 

Wc ¼    100%: ð11Þ The comparison of the wicking coefficient results of
WcðBÞ 
all samples by both methods are summarized in

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

522 Textile Research Journal 84(5)

Figure 13. Comparison of the results of the two measurements for cotton/polyester fabric (C/P-2) in the weft direction:
(a) relationship between the wicking length and time; (b) relationship between wicking length and square root of time.

Figure 14. Comparison of the results of the two measurements for cotton/polyester fabric (C/P-2) in the 45 bias direction:
(a) relationship between the wicking length and time; (b) relationship between wicking length and square root of time.

Figure 18. From this figure, it is evident that the wick-
ing coefficients determined for each fabric are very con-
sistent between the two measurement methods. It can
therefore be concluded that our thermocouple array
technique is suitable for tracking liquid movement
through fabrics. The data generated by the thermo-
couples also demonstrate the linear relationship
between wicking length and the square root of wetting
time predicted by the Washburn equation. Based on
these results, we are confident that the thermocouple

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

Zhu and Takatera 523

Figure 15. Comparison of the results of the two measurements for cotton knit fabric (C-k) in the warp direction: (a) relationship
between the wicking length and time; (b) relationship between wicking length and square root of time.

Figure 16. Comparison of the results of the two measurements for cotton knit fabric (C-k) in the weft direction: (a) relationship
between the wicking length and time; (b) relationship between wicking length and square root of time.

array technique can be used in place of the Byreck
method to follow the flow of water within fabrics.
Estimating water content from equilibrium
temperature of fabric
The relationship between the water content and the
temperature of each fabric samples during the drying
process is shown in Figure 19. For cotton woven fabric
(C), the fabric temperature exhibits very little change
until the water content approaches approximately 35%,
at which point the temperature increases significantly
with further decreases in the fabric’s water content.
These results reflect a typical drying process. In the
first period, the free water at the fabric surface evapor-
ates. The drying rate is determined by the diffusion of

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

524 Textile Research Journal 84(5)

Figure 17. Comparison of the results of the two measurements for cotton knit fabric (C-k) in the 45 bias direction: (a) relationship
between the wicking length and time; (b) relationship between wicking length and square root of time.

Table 2. Wicking coefficients (cm/s0.5) obtained for fabric

samples in three directions with two methods

Fabric Method Warp Weft 45 bias

C Thermocouple 0.3099 0.2433 0.3259

Byreck 0.3084 0.2493 0.3243
Wc (%) 0.4864 2.4067 0.4934
C/P-1 Thermocouple 0.4618 0.3202 0.4530
Byreck 0.4660 0.3021 0.4526
Wc (%) 0.9013 5.9914 0.0884
C/P-2 Thermocouple 0.5915 0.5618 0.5939
Byreck 0.5839 0.5391 0.5865
Wc (%) 1.3016 4.2107 1.2617
C-k Thermocouple 0.9733 0.7077 0.7333
Byreck 0.9171 0.6811 0.6993
Wc (%) 6.1280 3.9054 4.8620

Figure 18. Comparison of wicking coefficient results of differ-

water vapor from the saturated surface of the fabric
through a stationary air film into the air stream. This
is the constant rate period, which it will maintain for a
period of time. Eventually the water content of the
material drops to a level known as the critical water
content, and there is little free water, then water from
the inside of the fabric will transport to the surface and
evaporate, until the fabric is dry. This is the falling rate
period.27–29 In this cotton fabric, 35% water content is
evidently the critical water content, which refers to the
water content of the fabric at the end of the constant
ent fabric samples in three directions from the thermocouple and
Byreck methods.
period when the falling rate period begins. Above this
level, drying occurs in the constant rate period and
fabric temperature is approximately constant, whereas
below this level, drying enters the falling rate period
and the temperature increases greatly with decreasing
water content. The critical water content of cotton/
polyester fabric (C/P-1, C/P-2) is about 7%, and for
the cotton knit fabric, it is about 35%, too.

Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015

Zhu and Takatera
Figure 19. Relationship between water content and fabric
temperature during the drying process of all fabric samples.
For polyester knit fabric, the critical water content is
about 2%. The differences in critical water content are
due to the fiber content of fabric. The cotton fabrics (C
and C-k) have the highest critical water content, which
is caused by more water being absorbed by cotton
fabric, as well as yarn and fiber. It takes much time
for cotton yarn and fiber to release absorbed water
and therefore the critical water content is higher than
others.
Figure 20 presents results concerning the relation-
ship between the temperature difference and the water
content of cotton fabric at equilibrium. In this figure,
the data points labeled ‘‘Weight’’ represent temperature
difference calculated by subtracting fabric temperature
shown in Figure 19 from room temperature, while the
points labeled ‘‘C-1’’ to ‘‘C-6’’ represent the results of
thermocouple measurements from six replicate trials.
These results demonstrate that the measured tempera-
tures and weights are highly reproducible, and that
thermocouple results are consistent with the weight
results. At low values of temperature difference, the
slope of the plot between water content and tempera-
ture difference is quite small, whereas this slope
increases significantly at temperatures above approxi-
mately 3 C, corresponding to the water content of
about 35%.
Based on the analysis of critical water content,
Figure 20 can be divided into two parts, as shown by
the two regression lines. When the temperature differ-
ence reaches 3 C, the water content of fabric varies
greatly, and during this process, it is difficult to predict
water content only by temperature changes of fabric.
However, if the temperature difference is below 3 C, the
Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015
525
Figure 20. Comparison of results obtained by the weighing
method and thermocouple array method.
relationship between the water content and temperature
difference of each segment is highly in agreement with
the regression lines plotted by ‘‘Weight’’ in the experi-
mental trial. The relationship between temperature dif-
ference and water content of cotton fabric has been
verified. From the temperature difference, the dry
and wet conditions of fabric segment can be obtained
and, moreover, the water content of the fabric segment
can be estimated. We therefore conclude that the mea-
sured temperature difference between the fabric and
ambient can be used to estimate the water content of
the fabric, provided the water content of the fabric is
below the critical equilibrium water content for the
fabric.
Conclusions
A new method has been proposed for the measurement
of in-plane capillary water flow within fabrics. This
technique, using data acquired from a series of thermo-
couples, is capable of automatically tracking the wick-
ing length versus time. The wicking coefficients
obtained by this new method were compared with
those found using Byreck method, and the accuracy
of the thermocouple method was verified. As the meas-
urement thermocouples are physically small, removable
and flexible, this method can be applied to many differ-
ent kinds of textile structures. Moreover, this method
can be used to measure the wicking length with respect
to time without the addition of dye to the liquid.
In further trials, the relationship between tempera-
ture difference and water content was found based on
physically weighing fabric specimens, and the critical
526
water content was obtained. When the water content of
cotton fabric was above the critical water content of
about 35%, the temperature difference was essentially
constant. However, below this water content, the tem-
perature difference increased rapidly as the water con-
tent decreased. Based on these investigations, it appears
that it is possible to estimate the water content of fab-
rics from the thermocouple temperature measurements,
so long as the water content of the fabric is lower than
the critical value for that material. This method and the
associated apparatus are considered to be applicable to
many different areas of fabric analysis, including the
assessment of wetting in diapers and the dispersion of
sweat in sportswear.
Funding
This work was supported by Grants for Excellent Graduate
Schools, MEXT, Japan. This work was also supported by a
Grant-in-Aid for Global COE Program by the Ministry of
Education, Culture, Sports, Science, and Technology of
Japan, and by JSPS KAKENHI (grant number 24220012).
References
1. Erik K. Wetting and wicking. Textil Res J 1996; 66:
660–668.
2. Hsieh YL. Liquid transport in fabric structures. Textil
Res J 1995; 65: 299–307.
3. Rajagopalan D, Aneja AP and Marchal JM. Modeling
capillary flow in complex geometries. Textil Res J 2001;
71: 813–821.
4. Fangueiro R, Filgueiras A, Soutinho F, et al. Wicking
behavior and drying capability of functional knitted fab-
rics. Textil Res J 2010; 80: 1522–1530.
5. Patnaik A, Rengasamy RS, Kothari VK, et al. Wetting
and wicking in fibrous materials. Textil Prog 2006; 38:
1–105.
6. Kawase T, Sekoguchi S, Fujii T, et al. Spreading of
liquids in textile assemblies: Part I: Capillary spreading
of liquids. Textil Res J 1986; 56: 409–414.
7. Wang N, Zha AX and Wang JX. Study on the wicking
property of polyester filament yarns. Fibers Polym 2008;
9: 97–100.
8. Jiang XY, Zhou XH, Weng M, et al. Image processing
techniques and its application in water transport through
fabrics. J Fiber Bioeng Inform 2010; 3: 88–93.
9. Perwuelz A, Mondon P and Caze C. Experimental study
of capillary flow in yarns. Textil Res J 2000; 70: 333–339.
10. Lord PR. A comparison of the performance of open-end
and ring spun yarns in terry toweling. Textil Res J 1974;
44: 516–522.
11. Merve KO, Banu N and Cevza C. A study of wicking
properties of cotton-acrylic yarns and knitted fabrics.
Textil Res J 2011; 81: 324–328.
Downloaded from trj.sagepub.com at SONA COLG TECH on July 24, 2015
Textile Research Journal 84(5)
12. Esra C, Alex JF and Steven BW. Fibrous structures with
designed wicking properties. Textil Res J 2007; 77:
256–264.
13. Rossi RM, Stampfli R, Psikuta A, et al. Transplanar and
in-plane wicking effects in sock materials under pressure.
Textil Res J 2011; 81: 1549–1558.
14. Su CI, Fang JX, Chen XH, et al. Moisture absorption
and release of profiled polyester and cotton composite
knitted fabrics. Textil Res J 2007; 77: 764–769.
15. Japanese Industrial Standard L 1907:1994. Test method
of water absorbency of textiles-5.1.2, Byreck method.
16. Din 53924. Velocity of suction of textile fabrics in respect
of water- method determining the rising height. Berlin:
Deutsches Institut fur Normung, 1997.
17. Yoneda M and Niwa M. Measurement of in-plane capil-
lary water flow of fabrics. Sen’i Gakkaishi 1992; 48:
288–298.
18. Hsieh YL and Yu B. Liquid wetting, transport, and
retention properties of fibrous assemblies. Part I: water
wetting properties of woven fabrics and their constituent
single fibers. Textil Res J 1992; 62: 677–685.
19. Bayramli E and Powell RL. Experimental investigation of
the axial impregnation of oriented fiber bundles by capil-
lary forces. Colloids Surf 1991; 56: 83–100.
20. Ito H and Muraoka Y. Water transport along textile
fibers as measured by an electrical capacitance technique.
Textil Res J 1993; 63: 414–420.
21. Babu VR and Koushik CV. Capillary rise in woven fab-
rics by electrical principle. Indian J Fibre Textil Res 2011;
36: 99–102.
22. Zhu CH and Takatera M. Change of temperature of
cotton and polyester fabrics in wetting and drying pro-
cess. J Fiber Bioeng Inform 2012; 5: 433–446.
23. Mhetre S and Parachuru R. The effect of fabric structure
and yarn-to-yarn liquid migration on liquid transport in
fabrics. J Textil Inst 2010; 101: 621–626.
24. Washburn EW. The dynamic of capillary flow. Phys Rev
1921; 17: 273–283.
25. Kamath YK, Hornby SB, Weigmann HD, et al. Wicking
of spin finishes and related liquids into continuous fila-
ment yarns. Textil Res J 1994; 64: 33–40.
26. Nyoni AB. Liquid transport in Nylon 6.6 woven fabrics
used for outdoor performance clothing. In: Vassiliadis S
(ed.) Advances in modern woven fabrics technology.
Rijeka: InTech, 2011, pp.211–240.
27. Haghi AK. Thermal analysis of drying process: a theor-
etical approach. J Therm Anal Calorim 2003; 74: 827–842.
28. Richard S. Factors affecting the drying of apparel fabrics.
Part I: Drying behavior. Textil Res J 1958; 28: 136–144.
29. Sasaki S and Kubo J. Drying performance of cloths hung
in a thermostatic oven (Part 2). The relationship between
the drying rate and the surface temperature of wet cloths.
Nakumura Gakuen Univ 1980; 13: 129–133.